Evaluation of Hepatocellular Carcinoma With Dynamic $^{11}{\hbox {C}}$-Acetate PET: A Dual-Modeling Method

Abstract

Quantification of the <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">11</sup> C-acetate liver studies with dynamic positron emission tomography (PET) could significantly improve the evaluation of hepatocellular carcinoma (HCC), where both time-activity curves (TACs) of the hepatic artery (HA) and portal vein (PV) (the dual hepatic blood supply) are required. However, directly measuring them by the blood sampling or cannulation procedure is very invasive. In addition, it is very hard to differentiate the PV from the surrounding liver tissue on PET images by the currently developed indirect methods. To noninvasively and efficiently access the TAC of PV, we investigated the possibility of modeling the dual hepatic blood supply and presented two hepatic dual-input (DI) models. Combining the established <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">11</sup> C-acetate liver model with two different DI models, we obtained two dual-models with six/seven parameters to fit the dynamic PET measurements. The fitting results were compared with those of the <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">11</sup> C-acetate liver model using image-derived dual hepatic inputs (the "Golden standard") by statistical study. The adequacy of the two dual-models was estimated by Akaike Information Criteria (AIC) and Schwarz Criteria (SC). It was revealed that the proposed modeling technique could successfully account for the hepatic dual blood supply and the six-parameter (6-P) dual-model is more suitable for quantification of <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">11</sup> C-acetate liver studies.